How is the METRO variable (from the ACS) in the IPUMS database constructed?
Does it follow something like the following:
ACS is sent out to a random people.
Based on respondent’s location, IPUMS or ACS classifies whether ot not that respondent was in a metro area (based on metropolitan and micropolitan statstical areas in PUMAs) and labels 2,3, or 4 if they are.
If the METRO variable is not constructed this way, please share any information about how it is!
Thank you in advance.
The answer depends on the particular sample, but you are generally correct about how METRO is constructed. Details about this can be found on the Comparability Tab of the METRO variable. In all ACS samples, the lowest level geography is the PUMA and this information is used to assign metropolitan status. In 2005-2011 ACS samples, these PUMAs follow the boundaries from the 2000 Census. Therefore, in these years the METRO variable assigns metropolitan status based on classifications from 2000. In 2012 and onward ACS samples, these PUMAs follow the boundaries from the 2010 Census. Therefore, in these years the METRO variable assigns metropolitan status based on classifications from 2010.